"Before the birth of the torrent protocol in the early aughts, sharing big files, like TV shows or movies was virtually impossible"
This phrase is a bit misleading.
While not discounting the power of the torrent protocol, there were many p2p software before it which were more popular then it (napster, kazza, edunkey/eMule) .
The reason why sharing tv-shows and movies is so easy now then before is that we have broadband. In 2000 most people had dial-up.
If at all, the greatest achievement of the torrent protocol is allowing for discovery of files without a central repository.
The greatest achievement of the torrent protocol was the realization that giving up the order in which data is transfered a practical solution multicast problem.
Multicast was always talked about in networking previously, but solutions were found only for niche applications. These tended to have centralized setup requirements and other hurdles, so the internet remained basically entirely unicast.
I still find it incredible that a single source with a slow upload can broadcast to thousands of destinations - who add and remove themselves from the swarm freely, with newcomers automagically caught up - each receiving their full copy of the data mere seconds after the source has finished uploading a single copy.
In retrospect, older multicast ideas such as RFC 1112 (et al) were never going to work. Accomplishing the necessary routing while allowing dynamic group membership would have required adding an incredible amount of complexity switching fabric of the internet.
Another thing is that anyone could spin up a tracker.
Napster etc ran out of central corporate servers. This made them big juicy targets for lawyer bombing. Take out those servers and you take out the whole network.
But with bittorrent the tracker and the search are two separate entities. And over time you got things like DHT and multiple trackers for a single torrent that made it damn hard to knock them out.
Sony has resorted to flooding search sites with fake torrents in an attempt at slowing the spread of the files from the recent attack on their servers...
The BT protocol was a significant change in technology allowing files previously too large to share on traditional P2P networks to be shared.
Napster and the like were great for 3-4MB files, but anything larger would cause issues. You'd rely mostly on one-source, and if the connection to that source was shaky, the download would take days/weeks, even longer. BT changed this, and turned the idea of popular file=congested networks on it's ear completely. With BT, the more popular the file, the faster your download would go.
I used both Napster and Kazaa, both were entirely too slow to download large files on. I mean, you could, but you'd be waiting a long time.
I don't think it really had a whole lot to do with the protocols themselves though, but rather that broadband penetration in the States ramped up really quickly at about the same time Bittorrent came onto the scene. There was probably a short period of time where you could have downloaded TV shows on Kazaa, but that protocol died really fast.
Not many people used eDonkey back then. I tried. It was hard to find the stuff I wanted. Right after Kazaa died, both eDonkey and Bittorrent started taking off. Bittorrent just became the most popular protocol.
IRC, there where bots you could talk to. You could ask them what they got and download it. Worked like a charm, even for bigger files.
DDL, direct downloads, simple websites where you got links to download what you want.
eMule/eDonkey, those where running on my brothers machine day and night and rather good. There were links like bittorrent links back then. AFAIK you could start a DL pause it and switch to another hoster if the first one died.
I know there was a large piracy scene centered around AOL chat rooms. With AOL, once an attachment was uploaded to an email once, it could be forwarded endlessly without re-uploading the attachment.
Therefore there were dedicated "uppers" who would be given access to private scene FTP dumps. These uppers would create sequences of emails with files attached for various releases (usually individual rar or zip files).
They would forward the emails to people who ran bots in the chat rooms. You could request a list of files, and then a sequence of emails based by typing commands in chat. Since the files had already been uploaded, it was very fast for the chat bots to forward the emails as they were requested.
It was actually a pretty cool system. I remember calling AOL and giving them a story about how I needed to send email newsletters for my church, so could they please "whitelist" my account. Once that happened, you could send as much email as you wanted without being flagged.
I love this thread because it peels back layers of time. I'll go next: There was piracy on the BBSes! :) And let's not forget the binaries groups on Usenet.
I ran into a story about a guy in the UK that ended up involved in the piracy scene surrounding that platform.
The basic story was that in the UK there were a number of Amiga centric magazines printed. And in the back of them were several pages of classified ads. Amongst them were people offering to swap disks for disks via mail. You sent them a stack of disks and a list of what you wanted, and they would send you back what they had on that list.
So he put up a small ad, and got a few small envelopes. Over time this snowballed into him investing in multiple add-on drives and dedicating whole weekends to copying disks.
Indeed. In many cases, it would have been faster to drive across town and borrow the disk than to wait on a download. In my experience, with my 14.4k modem, downloading a 1 MB file took about an hour. If you're talking about long-distance BBS connections, unless you knew how to make free long-distance phone calls, it would be generally cheaper to buy the damn thing.
Human nature's funny. Societies have been fighting about the same things for a long time. My sister's friend used to burn and sell CDs and had pretty much all of the popular music from the time. If we go back further we can point at examples of 'piracy' brought about by the Gutenberg Press. (A beautiful machine by the way if you ever get a chance to see one operate.)
> My sister's friend used to burn and sell CDs and had pretty much all of the popular music from the time.
I think "piracy" is usually two different things: illicit copying for profit, and illicit copying for sharing. My first introduction was through sharing: cassette tapes for the vic20 and c64. Then floppies for the Amiga. Then BBSs (that where free to access, less the fee the phone companies took).
I think my first introduction to copying for (small) profit was around the time of the first affordable cd burners. Some people financed their cd burners this way -- and some made real money.
I never used Napster -- so I can't really comment. But with IRC and ftp sites -- things were again back to copying for sharing (no fee). Same for DC++/Direct Connect -- people ran hubs out of love, for fun -- and in many ways I'd say they were more distributed than torrent sites -- in the sense that there were many small (compared to the Pirate Bay) hubs, and there was more of a sense of community.
And again, no ads, no money involved.
I hope we'll see the rise of more distributed networks (eg: freenet) run by the users themselves, without any central orchestration -- and without an artificial ad-financed gateway like TPB. We'll see.
It's a shame Netflix can't just change to distributing torrents, as they'd never be allowed to license the content like that.
What I recall is when you used the IRC bots you were queued, sometimes behind hundreds of people. And then while downloading your or their connection would hiccup, and you would get re-queued.
I remember using IRC quite a bit, there were even sites that you could search and then go join the channel and talk to the bot for the ftp transfer etc. In hindsight FTP ratio requirements were sort of like the torrent ratio requirements some trackers implement.
I guess things were different here in Europe. I remember in High School (so... 2002ish? BitTorrent existed but there weren't any big popular trackers yet) We'd download a movie from Kazaa in the computer lab during CAD class and then watch during a long lunch break. Kazaa definitely supported using multiple sources.
Then when I went to Uni there was a large campus DC++ network with internal 100 MBit downloads.
> but rather that broadband penetration in the States ramped up really quickly at about the same time Bittorrent came onto the scene
Not only this, but prices for cheap virtual servers dropped like a rock, meaning that many more people could afford them or seedbox providers using them. Not many people running eMule or similar on a 100mbit or gigabit pipe.
I may be remembering incorrectly, but I think post-Napster clients like DC++ made it possible to resume a download from a different source, or pull the same file from multiple sources, by comparing checksums.
This made downloading popular files quite a bit faster, and eliminated most of the issues with p2p networks like Napster—it was just much less convenient because you had to search for the peers yourself.
Before bittorrent there was a very active pirate scene on usenet, irc and private ftp sites. It turns out that breaking up big files into smaller ones and being able to continue broken downloads will let just about anybody download just about anything even over dial-up speed connections.
Isn't that why RAR files were so popular for a while? Instead of just zipping up a 500MB file into a 350MB file for faster downloading, the .r01, .r02, etc. was perfect for downloading smaller chunks and only needing to find the bits you missed if something went wrong.
I have no idea if that's why it was done that way but it sure seemed to fit well. I'm pretty sure usenet is still a popular way to "pirate" TV shows and movies and such. It's just not as well known since it typically requires you to pay for access these days with most ISPs dropping support (and I doubt they'd allow access to the TV/movie groups anyway).
Originally the RAR file format was created to make it easy to put a large amount of data onto floppy disks, before writable CDs were widely available. The individual .r## files were sized to completely fill the disk space, so you didn't have to worry about organizing the files to minimize the number of disks needed.
After CD-RW drives became common RAR files were mostly used for uploading and downloading as you describe, to avoid file size caps and to make it somewhat easier to fix corrupted chunks. By the time the data being shared was filling the CDs we had writable DVDs, so RAR wasn't needed to span CDs. Then writable BluRay came around before DVD spanning became necessary.
Yeah, not only were they split up into multiple files they used a system to verify there is no corruption and if there is to reconstruct the corrupted files.
Torrent files brought the power of Usenet to the masses through ease of use and a total HTTP based solution (i.e. not having to pay extra for a separate Usenet account). I don't think connection speed had as much to do with anything, but the sheer number of new file sharers thrust torrents into the limelight.
The trouble always was being able to find all those missing pieces when the download took a long time. These scenes really encouraged a level of interaction with other users that is kind of missing today. Knowing someone who was consistently available and trading favors was somewhat common.
True, one thing though is that the Bittorrent protocol favours a 'tit for tat' type data exchange between peers unlike say, Emule which uses a queue system with a slight bonus for people who has given you data, this is in my view what made bittorrent quickly gain such widespread use as opposed to other existing p2p solutions.
I'd always thought eMule would have been much more successful if it answered the queue randomly, but weighting people who had uploaded with a much greater likelihood of going first.
As it was written eMule had a soft-cap of 10x "line speed" for someone who's uploaded 5x as much data to you as you have to them. This was far too little to get through a queue quickly. Worse, place in line wasn't remembered, so even a massive seeder on the eMule network might end up with zero files if he turned on the client for only half a day or so -- this is why eMule sharing typically consisted of having a computer run 24/7 and not expecting the file you wanted until a few days later.
Bittorrent changed that. While both clients were working with the same amount of underlying bandwidth, bittorrent made the connection between uploading and downloading much more immediate without an upper limit on tit-for-tat, and the incentive was clearly stronger.
Not really, thanks to magnet links and the dht you don't even need trackers. Sure, you need the magnet link, but you could even distribute them in grafitti or monthly paper issues if you felt inclined.
I was comparing it to Napster, and I thought Kazaa also had a central repository.
But both Kazaa and eMule have the same "supernode" like search mentality as bit-torrent (in bit-torrents it's the trackers).
I was referring more to the fact that unlike Kazaa/eMule where you needed to first find a "supernode" and then search for the file, bittorrent actually had the "tracker" url inside the download link (.torrent or magnet link).
>While not discounting the power of the torrent protocol, there were many p2p software before it which were more popular then it (napster, kazza, edunkey/eMule).
Even before those things became popular IRC was a popular way to trade files. The first MP3s that I downloaded came from an fserv on IRC. That lead me to alt.binaries.sounds.mp3.* on Usenet.
This phrase is a bit misleading.
While not discounting the power of the torrent protocol, there were many p2p software before it which were more popular then it (napster, kazza, edunkey/eMule) .
The reason why sharing tv-shows and movies is so easy now then before is that we have broadband. In 2000 most people had dial-up.
If at all, the greatest achievement of the torrent protocol is allowing for discovery of files without a central repository.